In my posting of Jun-25-2007, To Graph Or Not To Graph , I made the case (tentatively) that graphs weren't all they're cracked up to be, and provoked some lively discussion in the Comments section here. In his Apr-01-2009 posting, Why tables are really much better than graphs on the Statistical Modeling, Causal Inference, and Social Science Web log, Andrew Gelman makes a much more forceful case against graphs. Readers may find Gelman's arguments of interest.

I am not "anti-graph", but do think that graphs are often used when other tools (test statistics, tables, etc.) would have been a better choice, and graphs are certainly frequently misused. Thoughts?

Subscribe to:
Post Comments (Atom)

## 5 comments:

I work in Marketing & Advertising and can state with confidence that if you can't put the data in a chart or graph then you are better off not showing the data.

Uh....

Andy's posting was an April fool's joke.

He is hard-core pro-graph.

The problem with graphs is that there are no statistical tests to tell us that the graph is significant in of itself. If there were only a ... g-test or something....

I guess the eggs on my face, ha ha!

Still, there is still the question of how best to use graphs, and when graphs are simply not enough.

Comment from Gelman on his post: "As I hope was clear, in many ways my discussion above is serious."

Both tables and graphs can be misleading. If you report two parameter estimates, where one is higher than the other but the confidence intervals overlap - only a statistician will interpret this as not having enough evidence to judge the parameters to be unequal. Everyone else (meaning most people) will compare the two estimates and judge that one is higher than the other.

As always, representing data requires thought and a mixture of approaches.

Post a Comment